Segregation of Duties is one of those business concepts that’s a bit abstract, but the…
A lot of people assume that the built-in virtualisation and redundancy in cloud deployments automatically set you free from worrying about hardware. After all, that’s one of the primary reasons that companies choose virtualisation. There are no significant capital expenses to consider, and in most cases, you’re not really even running on a specific set of hardware — you’re operating in a cloud, where you can flexibly provision virtual machines any time you want.
For most deployments, if one of your VMs starts to fail, you spin up another one. If you run low on resources, you simply provision some more resources. If your cloud vendor is doing their job by replacing broken pieces and applying upgrades as needed, there’s nothing to worry about.
While that’s true for standard workloads, SAP HANA Hosting and some other high-performance workloads make it a bit more challenging.
How SAP HANA Cloud Solutions Are Different
That commonly held perception of cloud-based deployment holds true for web-scale workloads. These workloads are built to scale out on distributed server farms. With processing power and mass storage cheaply available, cloud providers can swap out pieces easily without any noticeable change in performance and zero downtime.
Because SAP HANA demands fast access to storage and RAM, SAP HANA hosting works differently than traditional hosting. Each HANA landscape is a customised solution, which needs to be carefully configured around the needs of the tenant. This has some major benefits. It provides the ability to build a landscape that fits both your needs and internal business logic exactly — something you can’t do with off-the-shelf, cloud-scale software. It can also be run to much higher performance and availability standards than typical web-scale workloads.
However, this also makes an SAP HANA cloud landscape more individualised, so there is less ability to just swap out parts, since they’re not cheap. You have high-performance storage as well as processors and large amounts of RAM that require additional planning and cost to provision or replace as quickly and easily as in a standard web-scale situation. Often the costs mean that problems need to be detected and repaired, rather than simply swapped out.
SAP HANA Hardware Level Insight
SAP HANA hosting in the cloud still provides a lot of the benefits of virtualisation. If standard hardware goes bad, you can move to new hardware, and you can scale when you start to run out of resources. The biggest difference is that it requires more out of your IT team than standard cloud-scale workloads. You need to monitor HANA from the hardware level up through the application level, so you can spot hardware faults and other issues before they become problems. And your team needs to provide daily maintenance, so that glitches don’t lead to service degradation, missed SLAs, and outages.
While both AWS and Azure come with basic monitoring packages that allow you to see most of your performance metrics, you also need specialists who can look at the entire infrastructure stack, from the OS level up. This is where a specialised cloud provider, such as Symmetry, which specialises in SAP HANA hosting, really makes a difference. Symmetry engineers have the monitoring capabilities for the entire infrastructure stack, so they can make sure that you are always meeting your performance and uptime SLAs.
Symmetry Has the HANA Hardware Experience
As a leader in enterprise application management and cloud hosting, we understand your whole technology stack, from the applications that run your business processes to the infrastructure and networking that your landscape is built on.
Whether you’re looking for a complete SAP HANA private cloud, hybrid-cloud solution, or someone to support your public-cloud landscape, our HANA hosting experience enables Symmetry to deliver the performance, reliability, and innovation your business depends on.
Read our free SAP HANA Deployment Guide, or contact us to learn more.